MDD is a common debilitating condition with over half of those diagnosed with MDD failing to achieve complete remission. New, more targeted, treatment paradigms are desperately needed. The development of more targeted treatments requires a better understanding of the psychopathology of MDD. The hypothalamic- pituitary-adrenal (HPA or stress) axis is possibly the most widely studied physiology pathway in MDD. This stems from the consistent finding that HPA axis hormones are elevated in many with MDD. Despite the amount of research being performed on this axis, a complete picture of the pathophysiology remains elusive. One reason for this lack of clarity is that the existing analysis methods fail to appropriately characterize the biologically important hormonal fluctuations in the HPA axis. The goal of this proposal is to improve the treatment options in MDD by removing the data analysis barriers in studies of the HPA axis and MDD. Studying the HPA axis is challenging because the hormones in the axis are secreted intermittently in boluses (called pulses), which are not directly observable. The axis is regulated by the pulse release patterns along with circadian rhythms. Current analyses of the pulse secretion are performed separately on each individual, and group comparisons are external to the model fitting. This results in unstable estimates of the pulse release features, especially a limited ability to characterize circadian rhythms, because the number of pulses on each subject is small. We address these shortcomings by developing a single model for a population of patients rather than a separate model for each patient. We use our new methods to more thoroughly investigate whether disease severity or classification is associated with: 1) the regularity of the pulsatile secretionor 2) changes in the circadian patterns in pulse frequency or pulse size. These are current mechanistic questions of interest in MDD studies of the HPA axis. We build upon our existing Bayesian approach to modeling pulsatile hormone data to overcome the computational issues that have precluded advancements in the analysis of hormones in the stress system. Our specific aims are: 1) to develop and evaluate a population model of pulsatile hormone data using a Bayesian approach, 2) to use these innovative methods to assess how ACTH and cortisol secretion differs between depressed and non-depressed participants, and 3) to disseminate the novel methodology and make it accessible to mental health researchers. Through this work, we will develop a novel analysis framework that investigators can use to better understand the role of the stress axis in psychopathology. In addition, this research lays a statistical foundation from which to build upon to characterize the complex feedback and forward loops that exist in the HPA axis. Ultimately, the work provides the mental health community with new tools critically useful for developing and evaluating new hormonally targets therapies for MDD and other mental health conditions.